650 EXPERIMENTAL INVESTIGATION OF THE 
introduced to the alloys whilst in a state of fusion, 
which had some probability in its favour, from the 
fact that the metal for silver coinage is fused in 
cast iron pots,* and, therefore, liable to take up 
a portion of those vessels. But, on the other 
hand, if that were always the practice, it would 
lead to the inference that in all the silver coins 
the iron would be nearly in the same proportion, 
and the extent of magnetic action almost the same 
in all. Whereas, by the tests already described, 
this is not the case. 
32. The current silver coinage of William and 
Mary became so base, that in the year 1694 it 
was all called in, and a new coinage issued. 
From this fact it occurred to me that there was a 
possibility, at least, that the high degree of 
magnetic action displayed by the half-crown of 
1691 (29), was owing to an undue proportion of 
copper, or of some other inferior metal. This 
idea led to the selection of a shilling, in which 
scarcely a trace of magnetism could be detected, 
for fusion with an additional portion of copper, 
also non-magnetic, having been obtained by the 
electro-type process. These, together with a 
* Ure’s Dictionary of Arts and Manufactures. Brande’s 
Chemistry, page 1037. 
